Automatic track defect positioning method and device
The automatic track defect location method combining online laser sensors and two-dimensional cameras solves the problem of difficult rapid location of track defects in existing technologies, and achieves efficient on-site defect location and analysis.
Patent Information
- Application Number
- CN202311574801.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-23
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-11-23
AI Technical Summary
Existing track inspection equipment cannot quickly and accurately locate the detected track defects, resulting in limited improvement in inspection efficiency. Furthermore, reliance on manual verification can easily lead to false positives and false negatives.
By employing online laser sensors and 2D cameras based on the starting profile features of track components, combined with mileage and sleeper information, track defects can be automatically identified and located at their specific positions on the track.
It enables accurate location of track defects, improves detection efficiency, reduces errors from manual verification, and ensures that inspectors can quickly locate defects for further analysis.
Smart Images

Figure CN117665844B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of track inspection technology, specifically relating to an automatic track defect location method and equipment. Background Technology
[0002] To ensure the safety and reliability of rail transit, regular inspections and maintenance of track infrastructure and basic equipment are necessary. This includes conducting multiple tests on various rail components, including the mainline rails and turnouts. Traditional inspection methods rely heavily on manpower. Inspectors manually inspect and identify various track defects on-site, using appropriate tools to measure their size, distribution, and other characteristics, and then record the data. This approach is clearly inefficient, leading to prolonged downtime for maintenance, and is prone to errors and omissions in manual inspections.
[0003] To improve inspection efficiency, some automated track inspection equipment and track defect detection methods have emerged. However, while these methods can automatically detect track defects to a certain extent, they cannot quickly locate the detected defects at their corresponding positions on the track. For example, existing technology includes a track inspection vehicle capable of data acquisition, automatic inspection of some items, detection of specific track defects, and obtaining their corresponding mileage data or satellite positioning data. For automatically detected defects, manual verification or further inspection and analysis are often still required. Mileage data or satellite positioning data are not intuitive location information for inspection personnel (natural persons), and the inspected track sections often contain multiple defects of different types. Therefore, inspection personnel cannot quickly locate the corresponding defects, resulting in limited improvement in inspection efficiency even with such automated inspection equipment and methods. Summary of the Invention
[0004] This invention addresses the aforementioned problems by providing an automatic track defect location method and equipment that pinpoints the location of detected track defects to the actual track location. This facilitates on-site verification or further analysis by inspection personnel. The invention employs the following technical solution:
[0005] This invention provides an automatic track defect location method, characterized by comprising: a detection start point location step: identifying the track start point from track profile data collected by an online laser sensor based on the track profile features of the track start point, obtaining corresponding mileage information, and obtaining the mileage information of the sleeper based on the mileage information and predetermined distance information between the track start point and nearby sleepers, as the mileage information of the detection start point; a track defect detection step: detecting track defects based on the track profile data and / or track images collected by a two-dimensional camera, the corresponding mileage information, and a predetermined track defect detection algorithm, and obtaining corresponding mileage information; and a track defect location step: obtaining at least one sleeper number corresponding to the track defect based on the mileage information of the detection start point, the mileage information of the track defect, and the sleeper information of multiple sleepers in the predetermined track.
[0006] The automatic track defect location method provided by this invention may also have the following technical features: it is used to locate track defects in turnout areas. Before the track defect location step, it further includes a detection direction determination step, which determines the current detection direction based on the type of the detected track component starting point, the track side corresponding to the track profile data of the detected track component starting point, and a predetermined detection direction determination rule. In the detection starting point location step, the type of the track component starting point is also obtained based on the profile features of the track component starting point. In the track defect location step, the sleeper number corresponding to the track defect is obtained based on the current detection direction, the mileage information of the detection starting point, the mileage information of the track defect, and the sleeper information.
[0007] The automatic track defect location method provided by this invention may also have the following technical features, wherein the sleeper information is turnout area sleeper information, including the sleeper number of each sleeper in the turnout area and the accumulated mileage corresponding to the current detection direction; the track defect location steps include: finding the corresponding accumulated mileage in the predetermined turnout area sleeper information based on the current detection direction and the type of the track component starting point; calculating the distance between the track defect and the detection starting point based on the mileage information of the detection starting point and the mileage information of the track defect; calculating the accumulated mileage corresponding to the track defect based on the accumulated mileage corresponding to the detection starting point and the calculated distance; finding the sleeper number and accumulated mileage of the corresponding target sleeper in the turnout area sleeper information based on the accumulated mileage of the track defect; and calculating the distance between the track defect and the target sleeper based on the accumulated mileage of the track defect and the accumulated mileage of the target sleeper.
[0008] The automatic track defect location method provided by the present invention may also have the following technical features, wherein the detection starting point location step includes: according to the mileage information and a predetermined track component starting point profile matching template, performing profile feature matching on multiple frames of track profile data in sequence to identify the track component starting point of the turnout track component and obtain its type and corresponding mileage information; based on the mileage information of the track component starting point, the predetermined distance between the track component starting point and the centerline of the nearest sleeper, obtaining the mileage information of the sleeper's centerline, as the mileage information of the detection starting point.
[0009] The automatic track defect location method provided by this invention may also have the following technical features: the type of the turnout area is a single turnout, having a straight track direction and a curved track direction; the turnout rail component includes at least a switch rail; the track profile data is acquired through a track detection device, which includes a first detection unit and a second detection unit, respectively used to acquire the track profile data on both sides of the track; the detection direction determination step includes: when the type of the rail component starting point is a switch rail starting point in a repulsive state, and the track profile data of the rail component starting point is detected from the first detection unit, the current detection direction is determined to be the straight track direction; when the type of the rail component starting point is a switch rail starting point in a close state, and the track profile data of the rail component starting point is detected from the first detection unit, the current detection direction is determined to be the curved track direction; in the track defect location step, the straight track accumulated mileage or the curved track accumulated mileage is correspondingly searched and acquired according to the current detection direction.
[0010] The automatic track defect location method provided by this invention may also have the following technical features: it is used to locate track defects on a mainline track, wherein the sleeper information is mainline sleeper information, including the sleeper spacing between two adjacent sleepers and the number of sleepers in each section; the sleeper number of the mainline track includes a section number and a sleeper number; the track defect location steps include: calculating the number of sleepers between the detection starting point and the track defect based on the mileage information of the detection starting point, the mileage information of the track defect, and the sleeper spacing; calculating the sleeper number of the target sleeper corresponding to the track defect based on the predetermined sleeper number of the detection starting point, the number of sleepers in the interval, and the number of sleepers in the section; and calculating the distance between the track defect and the target sleeper based on the mileage information of the detection starting point, the mileage information of the track defect, the sleeper spacing, and the number of sleepers in the interval.
[0011] The automatic track defect location method provided by the present invention may also have the following technical features, wherein the detection starting point location step includes: according to the mileage information and a predetermined track component starting point profile matching template, performing profile feature matching on multiple frames of track profile data in sequence to identify the track component starting point of the mainline rail and obtain its type and corresponding mileage information; based on the mileage information of the track component starting point, the predetermined distance between the track component starting point and the centerline of the nearest sleeper, obtaining the mileage information of the sleeper's centerline, as the mileage information of the detection starting point.
[0012] The automatic track defect location method provided by the present invention may also have the following technical features: in the track defect detection step, the corresponding track profile data is extracted based on the predetermined distance between the detection location and the detection starting point, and the profile in the extracted track profile data is compared with the standard profile at the predetermined detection location. When the deviation of the profile relative to the standard profile is greater than a predetermined deviation threshold, it is determined that the corresponding track defect has been identified, and the mileage information corresponding to the frame of track profile data is used as the mileage information of the track defect.
[0013] The automatic track defect location method provided by the present invention may also have the following technical features: in the track defect detection step, the track image is input into a trained neural network model for identifying specific track defects. The model outputs the location information of the identified track defects and converts the location information into the mileage information of the track defects based on the calibration relationship between the two-dimensional camera and the line laser sensor.
[0014] This invention provides an automatic track defect location device, characterized in that it comprises: a detection start point positioning unit, used to identify the track start point from track profile data collected by an online laser sensor based on the profile features of the track start point, obtain corresponding mileage information, and obtain the mileage information of the sleeper based on the mileage information and predetermined distance information between the track start point and nearby sleepers, as the mileage information of the detection start point; a track defect detection unit, used to detect track defects based on the track profile data and / or the track image and the mileage information, and obtain the corresponding mileage information; and a track defect location unit, used to obtain at least one sleeper number corresponding to the track defect based on the mileage information of the track defect, the mileage information of the detection start point, and a preset sleeper number of the sleeper at the detection start point.
[0015] Invention Function and Effect
[0016] According to the automatic track defect location method and equipment of the present invention, the method includes a detection start point location step, a track defect detection step, and a track defect location step. In the detection start point location step, the track component start point is identified based on its profile features, and the sleeper near the track component start point is located to obtain its mileage information. In the track defect detection step, track defects are detected and their mileage information is obtained. In the track defect location step, at least one sleeper number corresponding to the track defect is obtained based on the sleeper mileage at the detection start point and the mileage of the track defect. That is, the method can accurately locate the position of the detected track defect to a sleeper near the track on site. Based on the corresponding sleeper number, inspection personnel can quickly find the sleeper and the track defect near it on site, thereby facilitating the review, further inspection, and analysis of the automatically detected track defects, and improving the efficiency of on-site track inspection. Attached Figure Description
[0017] Figure 1 This is a flowchart of the automatic track defect location method and equipment in Embodiment 1 of the present invention;
[0018] Figure 2 This is a perspective view of the track inspection vehicle in Embodiment 1 of the present invention;
[0019] Figure 3 This is a schematic diagram of the sensor distribution in the vehicle body unit according to Embodiment 1 of the present invention;
[0020] Figure 4 This is a perspective view of the load-bearing wheel assembly in Embodiment 1 of the present invention;
[0021] Figure 5 This is a cross-sectional view of the load-bearing wheel assembly in Embodiment 1 of the present invention;
[0022] Figure 6 This is a side view of the clamping assembly in Embodiment 1 of the present invention;
[0023] Figure 7 This is a perspective view of the clamping component in Embodiment 1 of the present invention;
[0024] Figure 8 This is a flowchart of the starting point positioning step in Embodiment 1 of the present invention;
[0025] Figure 9 This is a schematic diagram of the profile of the point rail in the fitting state in Embodiment 1 of the present invention;
[0026] Figure 10 This is a schematic diagram of the profile cut of the point rail in the repulsion state in Embodiment 1 of the present invention;
[0027] Figure 11 This is a flowchart of the track defect detection steps in Embodiment 1 of the present invention;
[0028] Figure 12 This is a schematic diagram of the process for calculating the wear of the switch rail in Embodiment 1 of the present invention;
[0029] Figure 13 This is a flowchart of the track defect location steps in Embodiment 1 of the present invention;
[0030] Figure 14 This is a block diagram of the automatic track defect positioning device in Embodiment 1 of the present invention;
[0031] Figure 15 This is a schematic diagram of the structure of a single turnout in Embodiment 1 of the present invention;
[0032] Figure 16 This is a schematic diagram of the sleeper numbering of the mainline rail in Embodiment 2 of the present invention;
[0033] Figure 17 This is a flowchart of the track defect location steps in Embodiment 2 of the present invention.
[0034] Figure label:
[0035] Track inspection vehicle 100; vehicle body 20; vehicle body unit 21; housing 211; inclined end 2111a; 2D camera 215; refractor 216; first-line laser sensor 217a; second-line laser sensor 217b; third-line laser sensor 217c; fourth-line laser sensor 217d; bearing mechanism 30; bearing wheel assembly 31; bearing wheel bracket 311; bearing wheel shaft 312; bearing wheel 313; encoder receiving slot 3133; bearing 314; brake component 315; clamping assembly 32; clamping bracket 321; mounting block 3211; guide rod 3212; guide component 322; guide rail 3221; slider 3222 ; Elastic component 323; Wheel bracket 324; Pressure wheel 326; Locking component 327; Locking mating component 3271; Locking component 3272; Wrench component 328; Pushing mechanism 40; Automatic track defect positioning device 60; Sleeper information storage unit 601; Detection information storage unit 602; Positioning side communication unit 603; Detection starting point positioning unit 604; Detection direction determination unit 605; Track defect detection unit 606; Track defect positioning unit 607; Defect information storage unit 608; Positioning side control unit 609; Single turnout 90; Stock rail 91; Switch rail 92; Straight connecting rail 93; Guide curve rail 94; Guard rail 95; Wing rail 96; Center rail 97. Detailed Implementation
[0036] To make the technical means, creative features, objectives and effects of the present invention easy to understand, the automatic track defect location method and equipment of the present invention will be specifically described below in conjunction with embodiments and accompanying drawings.
[0037] <Example 1>
[0038] This embodiment provides an automatic track defect location method and equipment for detecting relevant track defects on track components in a turnout area, and automatically locating the detected track defects to the nearest sleeper. The turnout structure will be briefly described below.
[0039] Figure 15 This is a schematic diagram of the structure of a single turnout in this embodiment.
[0040] like Figure 15 As shown, this embodiment uses a single turnout as an example for specific explanation. The single turnout 90 has three openings A, B, and C, corresponding to four travel directions. Specifically, the first direction from opening A to opening B is the straight direction, along which the train travels straight from left to right through the turnout; the second direction from opening B to opening A is the straight direction, along which the train travels straight from right to left through the turnout; the first direction from opening A to opening C is the curved direction, along which the train switches at the turnout; the second direction from opening C to opening A is the curved direction, along which the train travels at the turnout and merges into the main track. This embodiment uses the first directions of the straight and curved tracks as examples for explanation.
[0041] The turnout rail components at point 90 of a single turnout include the stock rail 91, switch rail 92, straight connecting rail 93, guide curve rail 94, guard rail 95, wing rail 96, and frog rail 97 (turnout frog). Traveling in the first direction along the straight rail, the turnout starts at the switch rail 92 and passes the straight connecting rail 93; traveling in the first direction along the curved rail, the turnout starts at the switch rail 92 and passes the guide curve rail 94.
[0042] Figure 1 This is a flowchart of the automatic track defect location method in this embodiment.
[0043] like Figure 1 As shown, the automatic track defect location method includes the following steps:
[0044] Track data acquisition step S1: Acquire track profile data using a line laser sensor, acquire track images using a 2D camera, and obtain corresponding mileage information using a mileage encoder.
[0045] Step S2 for detecting the starting point: Based on the profile features of the starting point of the track component, identify the starting point of the track component in the track profile data, obtain the corresponding mileage information, and obtain the mileage information of the sleeper based on the mileage information and the distance information between the predetermined starting point of the track component and the nearby sleeper, which is used as the mileage information of the starting point of the detection.
[0046] Detection direction determination step S3: Based on the type of the detected track component starting point, the track side corresponding to the track profile data of the detected track component starting point, and the predetermined detection direction determination rules, determine the current detection direction.
[0047] Track defect detection step S4: Based on track profile data and / or track images, corresponding mileage information, and a predetermined track defect detection algorithm, track defects are detected, and mileage information of the track defects is obtained.
[0048] Track defect location step S5: Based on the current detection direction, the mileage information of the track defect, the mileage information of the detection starting point, and the sleeper number at the preset detection starting point, obtain the sleeper number corresponding to the track defect and the distance from the track defect to that sleeper.
[0049] The steps described above will be explained in detail below.
[0050] Track data acquisition step S1: Acquire track profile data using a line laser sensor, acquire track images using a 2D camera, and obtain corresponding mileage information using a mileage encoder.
[0051] As an example, the following shows a track inspection vehicle equipped with a wired laser sensor, a two-dimensional camera, and an odometer encoder for collecting the aforementioned data.
[0052] Figure 2 This is a perspective view of the track inspection vehicle in this embodiment.
[0053] like Figure 2 As shown, the track inspection vehicle 100 includes a vehicle body 20, a supporting mechanism 30, and a pushing mechanism 40. The vehicle body 20 is movably mounted on two mainline rails 9 via the supporting mechanism 30 at the bottom. The pushing mechanism 40 is installed on the vehicle body 20 and is used by inspection personnel to push the vehicle body 20.
[0054] The vehicle body 20 has a two-part structure, comprising two mirror-symmetrical vehicle body units 21. Each vehicle body unit 21 includes a housing 211 and sensors, electronic control components, etc., disposed within the housing 211.
[0055] Figure 3 This is a schematic diagram of the sensor distribution in the vehicle body unit in this embodiment.
[0056] like Figure 3As shown, each vehicle body unit 21 is equipped with a two-dimensional camera 215 and a refractor 216. The two-dimensional camera 215 is mounted horizontally on the top of the housing 211 via a corresponding bracket, with its lens facing horizontally. The refractor 216 is a 45° refractor, also mounted on the top of the housing 211 via a corresponding bracket, and positioned in front of the lens of the camera 215. Through the refraction of light by the refractor 216, the two-dimensional camera 215 can capture images of the rails below. This design makes the installation of the two-dimensional camera 215 more stable and avoids obstructing the line laser sensor.
[0057] Each car body unit 21 also has four line laser sensors inside its housing 211, which, for ease of description, will be referred to as the first line laser sensor 217a, the second line laser sensor 217b, the third line laser sensor 217c, and the fourth line laser sensor 217d, respectively. The housing 211 has an opening at the bottom, through which all four line laser sensors can detect the turnout rail components below.
[0058] The first line laser sensor 217a is mounted inside the housing 211 via a corresponding bracket and is located within the inclined end 2111a of the housing 211. The first line laser sensor 217a is mounted laterally. Taking the base rail as an example, when the track inspection vehicle 100 is placed on the base rail, the first line laser sensor 217a is located outside the base rail, slightly higher than the base rail, and facing the base rail. The line laser it projects covers the non-functional edge of the base rail, the lower jaw portion of the non-functional edge, and part of the rail surface, enabling it to acquire corresponding point data.
[0059] The second line laser sensor 217b is mounted at an angle on the top of the housing 211 via a corresponding bracket. When the track inspection vehicle 100 is placed on the track, the second line laser sensor 217b is located above and outside the main rail and facing the main rail. The line laser it projects covers the non-functional side, the waist of the non-functional side, and the rail surface of the main rail, and can acquire the corresponding point data.
[0060] The third-line laser sensor 217c is mounted obliquely at the top of the housing 211 via a corresponding bracket, and is closer to the center of the vehicle body 20 than the second-line laser sensor 217b. When the track inspection vehicle 100 is placed on the main rail, the third-line laser sensor 217c is located above and inside the main rail, facing the main rail. Its projected line laser covers the working edge, the waist of the working edge, and the rail surface of the main rail, enabling the acquisition of corresponding point data. When there are other turnout rail components inside the main rail, the line laser projected by the third-line laser sensor 217c can also cover or partially cover the upper end face and inner surface of the turnout rail component.
[0061] The fourth line laser sensor 217d is also mounted obliquely on the top of the housing 211 via a corresponding bracket. Its tilt angle is approximately the same as that of the third laser sensor 217c, and its installation position is closer to the middle of the vehicle body than that of the third laser sensor 217c. When the track inspection vehicle 100 is placed on the main rail, the fourth line laser sensor 217d is located on the inner side of the main rail, above and close to the space between the two main rails, and facing the bottom of the main rail. When there are other turnout rail components inside the main rail, the line laser projected by the fourth line laser sensor 217d covers the upper end face and inner side of the turnout rail component.
[0062] In this embodiment, all the line laser sensors on the track inspection vehicle 100 are of the same model. Furthermore, before the inspection begins, both the camera and the line laser sensors have been calibrated, and the corresponding calibration parameters have been obtained.
[0063] like Figure 2 As shown, the bearing mechanism 30 includes four bearing wheel assemblies 31 and four clamping assemblies 32. The four bearing wheel assemblies 31 are respectively installed on the lower parts of both sides of the vehicle body 20 in the width direction, and are installed in pairs near one end of the vehicle body 20 in the length direction, corresponding to the two rails 9 respectively. The four clamping assemblies 32 are also installed on the lower parts of both sides of the vehicle body 20 in the width direction, and are located next to the four bearing wheel assemblies 31. The track inspection device 100 is movably mounted on the two rails 9 via the four bearing wheel assemblies 31, at which time the four clamping assemblies 32 are located on the inner side of the corresponding side of the rail 9.
[0064] Figure 4 This is a perspective view of the load-bearing wheel assembly in this embodiment.
[0065] Figure 5 This is a cross-sectional view of the load-bearing wheel assembly in this embodiment.
[0066] like Figures 4-5 As shown, the bearing wheel assembly 31 includes a bearing wheel bracket 311, a bearing wheel shaft 312, a bearing wheel 313, a bearing 314, and a brake component 315.
[0067] The two ends of the bearing wheel shaft 312 are respectively mounted on the bearing wheel bracket 311 via two bearings 314. The bearing wheel 313 is mounted on the bearing wheel shaft 312 and is limited by the two bearings 314. In this embodiment, the bearing wheel 313 is made of insulating material, preferably ceramic or POM plastic, and the insulation resistance value of the bearing wheel 313 is not less than 1MΩ, to prevent the detection device 100 from connecting the two rails and conducting electricity, thus preventing interference with the rail's circuit system.
[0068] The bearing wheel 313 has an encoder receiving groove 3133 inside. Distance detection encoders 317 are installed in the encoder receiving grooves 3133 of two of the bearing wheels 313 to acquire corresponding mileage information as the bearing wheel 313 rolls along the rail 9. The size of the encoder receiving groove 3133 is slightly larger than the external dimensions of the distance detection encoder 317. Therefore, the distance detection encoder 317 does not directly contact the bearing wheel 313, avoiding friction that could affect the rotation of the bearing wheel 313.
[0069] It should be noted that the structure of the vehicle body 20 and the installation position of the line laser sensors enable multiple line laser sensors to scan the profile data of the same cross-section of the mainline rail 9. Under the control of the corresponding controller, during the movement of the trolley, multiple line laser sensors scan synchronously after each predetermined time interval or after moving a predetermined distance. Each frame of point data (point cloud) acquired by the line laser sensors contains hundreds to thousands of coordinate points, each containing mileage information (Y-axis), height information (Z-axis), width information (X-axis), and brightness information.
[0070] Figure 6 This is a side view of the clamping component in this embodiment.
[0071] Figure 7 This is a perspective view of the clamping component in this embodiment.
[0072] like Figures 6-7 As shown, the clamping assembly 32 includes a clamping bracket 321, a guide component 322, an elastic component 323, a wheel bracket 324, a clamping wheel shaft (not shown in the figure), a clamping wheel 326, a locking component 327, and a wrench component 328.
[0073] The clamping bracket 321 includes a mounting block 3211 and a guide rod 3212. The guide rod 3212 is a cylindrical rod that extends in the same direction as the length of the vehicle body 20.
[0074] The guide component 322 includes a guide rail 3221 and a slider 3222. The guide rail 3221 is fixedly mounted on the housing 211, and its extending direction is the same as that of the guide rod 3212. The slider 3222 is slidably mounted on the guide rail 3221.
[0075] The elastic element 323 is a spring, which is also fitted on the guide rod 3212. One end of the spring abuts against the mounting block 3211, and the other end is embedded in the cylindrical groove on the wheel bracket 324 and abuts against the bottom of the groove.
[0076] The clamping wheel shaft is installed in a notch at the end of the wheel body bracket 324, and the clamping wheel 326 is rotatably mounted on the clamping wheel shaft and partially embedded in the notch. The clamping wheel 326 is also made of the aforementioned insulating material.
[0077] Therefore, under the spring force of the elastic element 323, the wheel bracket 324 and its clamping wheel 326 can be pressed towards the inside of the rail, so that the clamping wheel 326 is tightly attached to the working edge of the rail. In this embodiment, when the rail inspection vehicle 100 is placed on the rail, the clamping wheel 326 is located 16cm below the rail surface.
[0078] The locking component 327 includes a locking mating part 3271 and a locking part 3272.
[0079] The locking element 3272 is a locking screw, which is installed on the mounting block 3211 at the corresponding locking hole, and the end of the screw can move along the hole.
[0080] The locking fitting 3271 is strip-shaped with multiple circular holes along its length. One end of the locking fitting 3271 is mounted on the wheel bracket 324, and the other end passes through the locking fitting hole on the mounting block 3211 and can move along the hole. The aforementioned locking fitting hole communicates with the locking fitting hole. When the screw end of the locking fitting 3272 extends downward, the screw end passes through the locking fitting hole to reach the locking fitting hole and then enters the circular hole on this end of the locking fitting 3271, thereby fixing (locking) the relative position of the wheel bracket 324 and the mounting block 3211.
[0081] In this embodiment, among the two clamping assemblies 32 corresponding to one of the rails 9, the locking member 3272 is in an unlocked state, and its elastic member 323 can move freely; in the two clamping assemblies 32 on the other side, the locking member 3272 is in a locked state, and the positions of the wheel bracket 324 and the mounting block 3211 are fixed, that is, the position of its clamping wheel 326 relative to the car body 20 is fixed. In other words, in this embodiment, the two clamping assemblies 32 on one side are in the form of fixed side wheels, and the two on the other side are in the form of spring side wheels.
[0082] The wrench assembly 328 includes two combined linkages for the inspector to adjust the relative position of the clamping wheel 326 with respect to the vehicle body 20, thereby making it easier to place the rail inspection vehicle 100 onto the rail.
[0083] The pushing mechanism 40 includes a trolley rod and a laptop stand mounted on the end of the trolley rod. The angle of the trolley rod is adjustable, and the angle of the laptop stand relative to the trolley rod is also adjustable.
[0084] In this embodiment, the inspection personnel place the track inspection vehicle 100 on the main rail at the turnout, and start the onboard sensors and data acquisition device. Then, they push the track inspection vehicle 100 along one of the four inspection directions. As the track inspection vehicle 100 moves along the inspection direction, it collects track profile data and corresponding mileage information.
[0085] In addition, in this embodiment, the line laser sensors in the two vehicle body units 21 are configured to acquire track profile data synchronously.
[0086] Step S2 for detecting the starting point: Based on the profile features of the starting point of the track component, identify the starting point of the track component in the track profile data, obtain the corresponding mileage information, and obtain the mileage information of the sleeper based on the mileage information and the distance information between the predetermined starting point of the track component and the nearby sleeper, which is used as the mileage information of the starting point of the detection.
[0087] In this embodiment, the method is used to detect profile defects in the switch rail and automatically locate the detected defects. Accordingly, the detection starting point is the sleeper starting point near the starting point of the switch rail.
[0088] Figure 8 This is a flowchart of the starting point location step in this embodiment.
[0089] like Figure 8 As shown, the detection starting point positioning step S2 specifically includes the following sub-steps:
[0090] Step S2-1: Based on the mileage information and the predetermined track component starting point profile matching template, perform profile feature matching on multiple frames of track profile data in sequence to identify the track component starting point and obtain its type and corresponding mileage information.
[0091] Specifically, in this embodiment, profile matching is performed sequentially on each frame of track profile data according to the profile matching template for the starting point of the switch rail. In each frame of track profile data, it is checked whether there is a profile within a predetermined position range. When a profile is determined to exist, the profile within that position range is extracted according to the template. Based on the extracted profile, the feature size is calculated, and it is then determined whether the calculated feature size matches the corresponding feature data in the template. If they match, it is determined that the starting point of the switch rail has been identified in the current profile data frame, and the mileage information corresponding to the profile data of that frame is used as the mileage information of the starting point of the switch rail.
[0092] Before the switch rail starting point, the measured profile data only includes the profile of the basic rail. At the switch rail starting point, the profile data further includes the profile of the switch rail. Since the relative positions of the basic rail and the switch rail are roughly fixed at the switch rail starting point, and the profile characteristics of the switch rail starting point are also known (with corresponding drawing data), the profile of the switch rail can be extracted from the measured track profile based on the relative positions of the basic rail and the switch rail. Furthermore, based on the profile characteristics of the switch rail starting point, it can be determined whether the extracted profile is the profile of the switch rail starting point.
[0093] Figure 9 This is a schematic diagram of the profile of the point rail in the fitting state in this embodiment.
[0094] like Figure 9As shown, taking a switch rail in a mating state as an example, the system determines whether a profile exists within a predetermined range of multiple positions in the profile data. When a profile is determined to exist within the first position range (corresponding to the position of the switch rail in a mating state), this portion of the profile is cropped based on a template. The profile lines of the inner rail web (roughly vertical straight lines) and the profile lines of the upper surface of the switch rail bottom (slightly inclined horizontal straight lines) are extracted. Then, it is determined whether these two straight lines exist within the extracted profiles. If yes, it is further determined whether the included angle between these two straight lines is a predetermined value. If yes, it indicates that the profile of the starting point of the switch rail in a mating state has been identified.
[0095] Figure 10 This is a schematic diagram of the profile cut of the point rail in the repulsion state in this embodiment.
[0096] like Figure 10 As shown, for a switch rail in a repulsion state, it is determined whether there is a profile within a predetermined range of multiple positions in the profile data. When it is determined that there is a profile within the second position range (corresponding to the switch rail in a repulsion state), the profile of that part is truncated and the feature size is calculated in the same way, thereby identifying the profile of the starting point of the switch rail in a repulsion state.
[0097] Step S2-2: Based on the mileage information of the starting point of the track component and the distance between the predetermined starting point of the track component and the nearest sleeper, obtain the mileage information of the sleeper, which is used as the mileage information of the detection starting point.
[0098] In this embodiment, the distance value from the starting point of the switch rail to the centerline (centerline in the width direction of the sleeper) of the sleeper closest to the starting point of the switch rail is pre-stored. The mileage of the centerline of the sleeper can be obtained by adding the mileage of the starting point of the switch rail to the distance value, which is used as the mileage of the detection starting point.
[0099] In this embodiment, information on all sleepers in the entire turnout area is pre-defined, and Table 1 below shows information on some of the sleepers as an example.
[0100] Table 1. Sleeper Information for Turnout Area
[0101]
[0102]
[0103] In Table 1, the units for the accumulated mileage of straight / curved tracks and the track spacing of straight / curved tracks are both mm. The sleeper spacing of straight / curved tracks also refers to the distance between the centerlines of two adjacent sleepers.
[0104] As shown in Table 1, the sleeper information for the predetermined turnout area includes sleeper information for straight tracks and sleeper information for curved tracks. The sleeper information for straight tracks includes the sleeper number along the straight track direction from the starting point of the entire turnout area, the distance between adjacent sleepers, the cumulative distance of the straight track to that sleeper, and the turnout position corresponding to the sleeper number (e.g., track start point, track inflection point, etc.). The sleeper information for curved tracks is similar and will not be described further. It can be seen that the distance between adjacent sleepers in the turnout area is not always equal, and the sleeper distances for straight and curved tracks also differ. Therefore, it is necessary to determine whether the current detection direction is along a straight or curved track.
[0105] Detection direction determination step S3: Based on the type of the detected track component starting point, the track side corresponding to the track profile data of the detected track component starting point, and the predetermined detection direction determination rules, determine the current detection direction.
[0106] In this embodiment, when the detected track component starting point is a switch rail starting point in a repulsion state, and the track profile data for detecting this switch rail starting point comes from the first vehicle body unit of the track inspection vehicle ( Figure 2 When the vehicle body unit 20 on the left side is detected, the current detection direction is determined to be the first direction of the straight track.
[0107] When the detected track component starting point is a contact point starting point and the track profile data of the detected point starting point comes from the first car body unit of the track inspection vehicle, the current detection direction is determined to be the first direction of the curved section.
[0108] It is understandable that, similarly, the detected track component starting point type is a contact point starting point, and the track profile data for detecting this point starting point comes from the second body unit of the track inspection vehicle. Figure 2 When the right-hand car body unit 20 is detected, the current detection direction is determined to be the first direction of the straight track. Alternatively, the detection results of the track profile data of the two car body units of the track inspection vehicle can be combined. That is, when the profile data of the first car body unit detects the starting point of the switch rail in a contact state and the profile data of the corresponding second car body unit (profile data of the same mileage) detects the starting point of the switch rail in a repulsive state, the current detection direction is determined to be the first direction of the straight track. If only one side of the car body unit detects it, corresponding warning information can be generated, which can further improve the accuracy of direction determination.
[0109] The second direction of the straight section and the second direction of the curved section can be determined similarly based on the profile of the detected guard rail starting point and which side of the car body unit the data for that profile comes from.
[0110] Track defect detection step S4: Based on track profile data and / or track images, corresponding mileage information, and a predetermined track defect detection algorithm, track defects are detected, and mileage information of the track defects is obtained.
[0111] In this embodiment, based on the track profile data of the switch rail and a predetermined switch rail profile detection algorithm, the profile defects of the switch rail are detected and the wear of the switch rail is measured.
[0112] Figure 11 This is a flowchart of the track defect detection steps in this embodiment.
[0113] like Figure 11 As shown, track defect detection step S4 specifically includes the following sub-steps:
[0114] Step S4-1: Based on the mileage information of the starting point of the switch rail and the distance information between the predetermined position to be detected and the starting point of the switch rail, extract the data of the position to be detected from the multi-frame track profile data.
[0115] Step S4-2: Preprocess the extracted track profile data to obtain the target profile.
[0116] The preprocessing includes filtering the track profile data to initially remove certain unwanted edges, and connecting the breakpoints of the profile to form a complete profile based on a predetermined algorithm.
[0117] Step S4-3: Using the jaw portion on the non-functional side of the basic track as a reference, match the target profile with the standard basic track profile.
[0118] The non-functional side of the basic rail does not contact the wheel, so it is basically free from wear and is not easily obscured by dirt. Furthermore, the lower jaw of the non-functional side has an easily identifiable corner. Therefore, using this lower jaw as the matching reference can improve the matching accuracy, and the matching algorithm is relatively easier to implement.
[0119] Step S4-4: Based on the profile features of the switch rail and the relative positions of the base rail and the switch rail, select the switch rail profile from the matched target profiles.
[0120] Step S4-5: Using the inner rail web and the upper end face of the rail bottom as a reference, match the selected switch rail profile with the standard full-section profile of the switch rail.
[0121] Step S4-6: Calculate the first deviation of the switch rail profile relative to the matching standard full-section profile, and calculate the second deviation of the standard switch rail profile at the test position relative to the standard full-section profile. Calculate the difference between the first and second deviations as the wear amount of the switch rail at the test position.
[0122] The standard switch rail profile is the ideal, wear-free switch rail profile at the location to be tested. The standard full-section profile is the profile of the rear portion of the switch rail, which is similar to the basic rail profile.
[0123] Step S4-7: Determine whether the wear of the switch rail at the location to be detected is greater than the predetermined wear threshold. If the determination is yes, the corresponding track defect (switch rail profile wear) is identified, and the mileage information corresponding to the profile data of that frame is used as the mileage information of the track defect.
[0124] Figure 12 This is a schematic diagram of the process for calculating the wear of the switch rail in this embodiment.
[0125] like Figure 12 As shown, the above steps S4-6 specifically include the following sub-steps:
[0126] Step S4-6-1: Based on the mileage information of the location to be detected, determine whether the location to be detected is one of multiple predetermined key locations.
[0127] Step S4-6-2: If the determination in step S4-6-1 is yes, extract the corresponding one from multiple predetermined key switch profiles based on the mileage information.
[0128] Step S4-6-3: Using the inner rail web and the upper end face of the rail bottom as a reference, match the extracted key switch rail profile with the standard full-section profile, and calculate the first deviation of the key switch rail profile relative to the matched standard full-section profile.
[0129] The first deviation includes the vertical distance between the apex of the critical switch profile and the apex of the standard full-section profile, and the horizontal distance between the side wear detection points on the working edge of the critical switch profile and the working edge of the standard full-section profile.
[0130] Since there are corresponding drawing data for specific locations on the switch rail, when the location to be inspected happens to be one of these specific locations, the corresponding standard profile can be obtained from the drawing data to calculate the wear amount.
[0131] Step S4-6-4: If the determination in step S4-6-1 is negative, extract the key switch rail profile of the key position adjacent to the position to be detected based on the mileage information, and determine the switch rail interval to which the position to be detected belongs based on the mileage information, and obtain the corresponding interval gradient value.
[0132] Step S4-6-5: Using the inner rail web and the upper end face of the rail base as a reference, match the extracted key switch rail profile with the standard full-section profile, calculate the first deviation of the key switch rail profile relative to the matched standard full-section profile, and correct the deviation based on mileage information and interval gradient value to obtain the first deviation.
[0133] Among them, the switch rail is a variable cross-section rail component. There is no corresponding drawing data for non-critical positions, but its cross-sectional shape changes gradually in a regular manner. Therefore, it can be calculated based on the profile of the closest critical position, the distance between the position to be detected and the closest critical position, and the interval gradient value.
[0134] Step S4-6-6: Using the inner rail web and the upper end face of the rail bottom as a reference, match the measured rail profile with the standard full-section profile, and calculate the second deviation of the rail profile relative to the matched standard full-section profile.
[0135] The second deviation also includes the vertical distance between the two vertices and the horizontal distance between the two working side grinding detection points.
[0136] Step S4-6-7: Calculate the difference between the first deviation and the second deviation, which is taken as the wear amount of the switch rail at the position to be tested.
[0137] Specifically, the difference between the two vertical spacing values in the first and second deviations is calculated as the vertical wear value of the switch rail at the position to be tested; the difference between the two horizontal spacing values in the first and second deviations is calculated as its side wear value.
[0138] Using the above method, even at the detection location where there is no standard drawing data, the accurate wear value of the switch rail can be obtained; and the wear value of each position of the switch rail can be detected using the same algorithm, without the need to configure multiple different matching algorithms and wear calculation algorithms due to changes in the switch rail cross section.
[0139] The above is only an example to illustrate one type of track defect detection method. In addition, another type of track defect can be detected based on track images acquired by a camera. For example, after preprocessing the acquired track images, the images are input into a trained neural network model to identify the blocks falling on the track surface, obtain the anchor frame coordinates of the blocks, and convert the anchor frame coordinates into track profile coordinates according to the calibration relationship between the camera and the line laser sensor, thereby obtaining the mileage information of the blocks falling on the track surface.
[0140] Track defect location step S5: Based on the detection direction, track defect mileage information, detection start point mileage information, and preset sleeper number at the detection start point, obtain the sleeper number of the target sleeper corresponding to the track defect, as well as the distance information from the track defect to the target sleeper.
[0141] Figure 13 This is a flowchart of the track defect location steps in this embodiment.
[0142] like Figure 13 As shown, the track defect location step S5 specifically includes the following sub-steps:
[0143] Step S5-1: Based on the detection direction and the type of the starting point of the rail component, find the corresponding sleeper number and the corresponding accumulated mileage in the sleeper information of the predetermined turnout area.
[0144] For example, if the current detection direction is the first direction of the straight track and the type of the starting point of the rail component is the switch rail starting point, then the sleeper number corresponding to the switch rail starting point can be found in Table 1 as straight track 8. The cumulative mileage of the straight track corresponding to this sleeper is 4500. The mileage unit is mm, and the same applies below.
[0145] Step S5-2: Calculate the distance between the track defect and the detection starting point based on the mileage of the detection starting point and the mileage of the track defect.
[0146] Step S5-3: Based on the accumulated mileage corresponding to the detection starting point and the calculated spacing, obtain the accumulated mileage corresponding to the track defects.
[0147] For example, if the distance between the detection starting point and the defect is calculated to be 3500, then the accumulated distance corresponding to the track defect is 4500 + 3500 = 8000.
[0148] Step S5-4: Based on the accumulated mileage of track defects, find the corresponding target sleeper's sleeper number and accumulated mileage in the predetermined turnout area sleeper information. Among all the sleepers with accumulated mileage less than that of track defects, find the one with the closest accumulated mileage of track defects as the target sleeper and obtain the sleeper number of the target sleeper.
[0149] For example, if the accumulated mileage corresponding to track defects is 8000, among all sleepers with a straight track accumulated mileage of less than 8000, the one with the closest accumulated mileage to 8000 is found. The target sleeper with a straight track accumulated mileage of 7500 is found, and its sleeper number is straight track 13.
[0150] Step S5-5: Based on the accumulated mileage of track defects and the accumulated mileage of the target sleeper, calculate the distance between the track defects and the target sleeper.
[0151] For example, if the accumulated mileage of a straight track defect is 8000 meters and the accumulated mileage of the target sleeper is 7500 meters, then the distance between the track defect and the target sleeper is 8000 - 7500 = 500 meters. This means the track defect is accurately located 500 mm away from sleeper numbered Straight Track 13. At the track site, each sleeper, or every few sleepers, is marked with its corresponding sleeper number. Therefore, inspectors can easily and intuitively locate the sleeper and quickly find the location of nearby track defects based on the aforementioned distance.
[0152] The above explanation uses straight stocks as an example. Curved stocks are calculated in the same way and will not be repeated here.
[0153] Afterwards, the detected track defects type, defect characteristic data (such as the aforementioned vertical wear value and lateral wear value), the corresponding profile data frame / 2D image frame of the track defect, the sleeper number of the target sleeper, and the distance between the track defect and the target sleeper can be stored for inspection personnel to query and further analyze. Through the sleeper number of the target sleeper and the distance information between the defect and the target sleeper, the inspection personnel can easily locate the corresponding position on the track site so that they can go to that position for manual verification or further inspection.
[0154] Figure 14 This is a block diagram of the automatic track defect location device in this embodiment.
[0155] like Figure 14 As shown, this embodiment also provides an automatic track defect location device 60 corresponding to the above method. The device includes a sleeper information storage unit 601, a detection information storage unit 602, a positioning side communication unit 603, a detection starting point positioning unit 604, a detection direction determination unit 605, a track defect detection unit 606, a track defect location unit 607, a defect information storage unit 608, and a positioning side control unit 609.
[0156] The sleeper information storage unit 601 stores sleeper information for the turnout area as shown in Table 1. The detection information storage unit 602 stores parameters and other information required for defect detection and location. The positioning side communication unit 603 is used to communicate with other devices, including acquiring track profile data, track images, and corresponding mileage information from the data acquisition device on the track inspection vehicle. The detection start point positioning unit 604 positions the sleeper at the detection start point according to the method described in step S2 above. The detection direction determination unit 605 automatically determines the current detection direction according to the method described in step S3 above. The track defect detection unit 606 detects track defects according to the method described in step S4 above. The track defect location unit 607 locates track defects according to the method described in step S5 above. The defect information storage unit 608 stores the above track defect information, including the sleeper number and distance information obtained through location. The positioning side control unit 609 controls the operation of the above functional units.
[0157] As one example, the automatic track defect location device 60 is a laptop computer with corresponding computer programs installed. This laptop computer is mounted on a laptop holder on the track inspection vehicle 100 and connected to the on-board data acquisition device via a cable. Inspection personnel can conveniently obtain the location results of track defects using this laptop computer.
[0158] In this embodiment, the parts not described in detail are well-known technologies in the art.
[0159] Functions and effects of Example 1
[0160] According to the automatic track defect location method and equipment provided in this embodiment, the method includes a detection start point location step, a track defect detection step, and a track defect location step. In the detection start point location step, the starting point of the track component is identified based on its profile features, and the sleeper near the starting point is located to obtain its mileage information. In the track defect detection step, track defects are detected and their mileage information is obtained. In the track defect location step, at least one sleeper number corresponding to the track defect is obtained based on the sleeper mileage at the detection start point and the mileage of the track defect. That is, this method can accurately locate the detected track defect to a sleeper near the track on-site. Based on the corresponding sleeper number, inspection personnel can quickly find the sleeper and the track defect near it on-site, thereby facilitating the review, further inspection, and analysis of the automatically detected track defects, and improving the efficiency of on-site track inspection.
[0161] In this embodiment, track profile data, two-dimensional images, and mileage information are collected via a line laser sensor, camera, and mileage encoder mounted on the track inspection vehicle. Because the track inspection vehicle uses wide bearing wheels, even with wear on the working edge of the rail, the bearing wheels can still roll stably along the surface of the rail, driving the internal mileage encoder to measure the mileage. The mileage encoder is located in a groove inside the bearing wheel with a certain circumferential distance from the groove, so the mileage encoder does not affect the rotation of the bearing wheel. The track inspection vehicle is also equipped with four clamping components, thus enabling very accurate mileage data measurement. On the other hand, the sleepers are fixed to the base structure, and their positions are not easily changed. Therefore, combining the measured mileage data and sleeper information allows for accurate location of track defects. Compared to the entire track section being inspected, the sleepers are densely distributed and spaced relatively close together; therefore, by locating the track defect to the corresponding sleeper number, inspection personnel can quickly find the defect location on-site.
[0162] In this embodiment, the method is used to locate defects in turnout areas. Since turnouts are divided into straight and curved tracks, and the distribution spacing of sleepers in straight and curved tracks is different, the current detection direction is automatically identified and determined based on the detected track component type and the car body unit corresponding to the detected track component profile data. This determines whether the current track is straight or curved, and then the sleeper number near the defect is calculated accordingly. In this way, the inspection personnel do not need to set the inspection direction in advance, making the operation more convenient and faster, and avoiding errors caused by manual setting.
[0163] In this embodiment, not only is the sleeper number of the sleeper adjacent to the track defect calculated based on the accumulated mileage, but the distance from the defect location to the center line of the sleeper is also calculated, making the positioning more accurate. When there are multiple detected track defects between two adjacent sleepers, one of them can be located conveniently and quickly based on this distance.
[0164] In this embodiment, the detected track defect types, characteristic size data, original collected data, corresponding sleeper numbers and distances are also stored to facilitate inspection personnel to query and conduct further analysis. Subsequently, based on the stored information, a visual distribution result of the defects can be generated to provide inspection personnel with more intuitive and easy-to-understand analytical results.
[0165] <Example 2>
[0166] This embodiment provides an automatic track defect location method and device. Compared with Embodiment 1, the difference is that the method in this embodiment is used to automatically locate track defects on the mainline rail.
[0167] In the track data acquisition step S1, the inspection personnel place the track inspection vehicle on the main line rail, set the sleeper number at the starting point of the inspection, and push the track inspection vehicle along the main line rail to collect data.
[0168] The process involves placing the track inspection vehicle in front of the rail gap between two adjacent rails. The vehicle is then moved across the gap. The collected profile data at this rail gap does not show the profile of the mainline rail. However, at the starting point of a section of mainline rail after this gap, the profile of the mainline rail appears in the data frame. This allows for the identification of the starting point of the mainline rail, and the sleeper closest to this starting point can be used as the inspection starting point.
[0169] In the detection starting point positioning step S2, similarly, based on the mileage information and the predetermined mainline rail starting point profile matching template, the profile feature matching of multiple frames of track profile data is performed sequentially to identify the starting point of the mainline rail and obtain the corresponding mileage information; then, based on the distance information between the starting point of the mainline rail and the sleeper closest to it, the mileage information of the sleeper is obtained, which is used as the mileage information of the detection starting point.
[0170] In track defect detection step S3, similarly, profile matching can be performed based on the measured track profile data and the standard profile of the mainline rail to detect the wear of the mainline rail; or the surface damage such as rail surface spalling of the mainline rail can be detected based on the measured track image. After the corresponding track defects are detected, the corresponding mileage information is obtained, which will not be elaborated further.
[0171] Since mainline rails do not have straight or curved strands like turnouts, step S4 can be omitted.
[0172] Figure 16 This is a schematic diagram of the sleeper numbering of the mainline rail in this embodiment.
[0173] like Figure 16As shown, in this embodiment, the sleeper number of the mainline rail includes the mileage range, CPIII number, section number, and sleeper number. For example, a sleeper number of 1534311002 represents the second sleeper in the 11th section of the left track (odd CPIII numbers represent the left track) within the mileage range of K1534. The spacing between two adjacent sleepers is basically a fixed value. Therefore, based on the sleeper number of the sleeper at the detection starting point and the distance from the detection starting point to the location of the track defect, the sleeper number near the track defect location can be calculated.
[0174] Figure 17 This is a flowchart of the track defect location steps in this embodiment.
[0175] like Figure 17 As shown, the track defect location step S5 specifically includes the following steps:
[0176] Step S5-1: Based on the mileage M0 of the detection starting point, the mileage M1 of the track defect, and the predetermined spacing d between two adjacent sleepers, calculate the number of sleepers n1 between the detection starting point and the track defect.
[0177]
[0178] Step S5-2: Based on the sleeper number N0, the number of sleepers n1, and the predetermined number of sleepers k for each section at the detection starting point, calculate the sleeper number N1 of the target sleeper adjacent to the track defect:
[0179]
[0180] In this embodiment, detection and corresponding calculations are performed along the forward direction of the track; that is, along the detection direction, the interval number increases, and the sleeper number within each interval also increases. If detection is performed in the reverse direction, the formula can be adjusted accordingly.
[0181] Step S5-3: Based on the mileage M0 of the detection starting point, the mileage M1 of the track defect, the predetermined spacing d between two adjacent sleepers, and the number of sleepers n1, calculate the distance D1 between the track defect and the target sleeper:
[0182] D1 = M1 - M0 - (d*n1)
[0183] It is understood that in this embodiment, as in Embodiment 1, the sleeper information of the mainline rail can be pre-stored, including the sleeper number of each sleeper, the spacing between two adjacent sleepers, the cumulative mileage from a predetermined starting point of the mainline rail to each sleeper, etc., and the sleeper number of the sleeper adjacent to the location of the track defect can be obtained using a method similar to that in Embodiment 1.
[0184] In this embodiment, the structure of the track inspection vehicle and other steps are the same as in Embodiment 1, so they will not be described again.
[0185] Functions and effects of Example 2
[0186] The automatic track defect location method and equipment provided in this embodiment are used to automatically locate defects in the main line track. Based on the characteristics of the main line track detection section being relatively long and its sleeper spacing being evenly distributed, the sleeper number and distance corresponding to the track defect are calculated based on the sleeper number of the detection starting point, the sleeper spacing, the mileage of the detection starting point, and the mileage of the track defect. This method can accurately locate the detected track defect in the vicinity of the corresponding sleeper, making it convenient for inspection personnel to find the defect on-site. Moreover, using this calculation method, it is not necessary to pre-store a large amount of sleeper information.
[0187] The above embodiments are only used to illustrate specific implementations of the present invention, and the present invention is not limited to the scope of the description of the above embodiments.
[0188] In the above embodiments, as an example, the track defect detection step includes the detection of track component profile and the detection of track surface blockage. In an alternative, other existing defect detection methods based on track profile data and / or track images can also be used in the track defect detection step to perform the corresponding defect detection.
[0189] In the above embodiment, the track defect is located near the sleeper closest to it along the detection direction, and the distance between the defect and the sleeper is provided, so as to accurately locate the defect on site. In an alternative, the track defect can also be located between two adjacent sleepers. Since the sleepers are densely distributed relative to the overall length of the track section, this also allows the inspection personnel to find the defect location relatively quickly.
Claims
1. A method for automatically locating track defects, characterized in that, include: Starting point detection steps: Based on the profile features of the starting point of the track component, the starting point of the track component is identified from the track profile data collected by the online laser sensor, and the corresponding mileage information is obtained. Based on the mileage information and the predetermined distance information between the starting point of the track component and the nearby sleepers, the mileage information of the sleepers is obtained as the mileage information of the starting point of the detection. The type of the starting point of the track component is obtained based on the profile features of the starting point of the track component. Track defect detection steps: Based on the track profile data or track images acquired by a two-dimensional camera, the corresponding mileage information, and a predetermined track defect detection algorithm, track defects are detected, and the corresponding mileage information is obtained; The detection direction determination step determines the current detection direction based on the type of the detected track component starting point, the track side corresponding to the track profile data of the detected track component starting point, and the predetermined detection direction determination rules. Track defect location steps: Based on the current detection direction, the mileage information of the detection starting point, the mileage information of the track defect, and the sleeper information of multiple sleepers in the predetermined track, at least one sleeper number corresponding to the track defect is obtained. Wherein, when used to locate track defects in a turnout area, the sleeper information is turnout area sleeper information, which includes the sleeper number of each sleeper in the turnout area and the accumulated mileage along the current detection direction. The track defect location steps include: Based on the current detection direction and the type of the starting point of the rail component, the corresponding accumulated mileage is obtained by searching the sleeper information in the predetermined turnout area; The distance between the track defect and the detection starting point is calculated based on the mileage information of the detection starting point and the mileage information of the track defect. Based on the accumulated mileage corresponding to the detection starting point and the calculated spacing, the accumulated mileage corresponding to the track defect is calculated; Based on the accumulated mileage of the track defects, the sleeper number and accumulated mileage of the corresponding target sleeper are found in the sleeper information of the turnout area. Based on the accumulated mileage of the track defects and the accumulated mileage of the target sleeper, the distance between the track defects and the target sleeper is calculated. When used to locate track defects on the mainline track, the sleeper information is mainline sleeper information, which includes the sleeper spacing between two adjacent sleepers and the number of sleepers in each section. The sleeper number of the mainline track includes the section number and the sleeper number. The track defect location steps include: Based on the mileage information of the detection starting point, the mileage information of the track defect, and the sleeper spacing, calculate the number of sleepers between the detection starting point and the track defect. Based on the sleeper number of the predetermined detection starting point, the number of sleepers at intervals, and the number of sleepers in the section, calculate the sleeper number of the target sleeper corresponding to the track defect; Based on the mileage information of the detection starting point, the mileage information of the track defect, the sleeper spacing, and the number of spaced sleepers, the distance between the track defect and the target sleeper is calculated.
2. The automatic track defect location method according to claim 1, Its features are: The detection start point identification step includes: Based on the mileage information and the predetermined track component starting profile matching template, the profile feature matching is performed on multiple frames of track profile data in sequence to identify the track component starting point of the turnout track component and obtain its type and corresponding mileage information. Based on the mileage information of the starting point of the rail component, the predetermined distance between the starting point of the rail component and the centerline of the nearest sleeper, the mileage information of the centerline of the sleeper is obtained, which is used as the mileage information of the detection starting point.
3. The automatic track defect location method according to claim 2, Its features are: The type of turnout area is a single turnout, which has a straight track direction and a curved track direction. The turnout rail components include at least a switch rail. The track profile data is acquired through a track detection device, which includes a first detection unit and a second detection unit, respectively used to acquire the track profile data on both sides of the track. The detection direction determination step includes: When the type of the starting point of the track component is detected as a point rail starting point in a repulsion state, and the track profile data of the starting point of the track component is detected as coming from the first detection unit, it is determined that the current detection direction is the straight direction; When the type of the starting point of the track component is detected as a point rail starting point in a mating state, and the track profile data of the starting point of the track component is detected as coming from the first detection unit, it is determined that the current detection direction is the curved direction. In the track defect location step, the straight track accumulated mileage or the curved track accumulated mileage is located and obtained according to the current detection direction.
4. The automatic track defect location method according to claim 1, Its features are: The detection start point identification step includes: Based on the mileage information and the predetermined track component starting point profile matching template, the profile feature matching is performed on multiple frames of track profile data in sequence to identify the starting point of the track component of the mainline rail and obtain its type and corresponding mileage information. Based on the mileage information of the starting point of the rail component, the predetermined distance between the starting point of the rail component and the centerline of the nearest sleeper, the mileage information of the centerline of the sleeper is obtained, which is used as the mileage information of the detection starting point.
5. The automatic track defect location method according to claim 1, characterized in that: in, In the track defect detection step, the corresponding track profile data is extracted based on the predetermined distance between the detection location and the detection starting point. The profile in the extracted track profile data is compared with the standard profile at the predetermined detection location. When the deviation of the profile relative to the standard profile is greater than a predetermined deviation threshold, the corresponding track defect is identified, and the mileage information corresponding to the track profile data is used as the mileage information of the track defect.
6. The automatic track defect location method according to claim 1, characterized in that: in, In the track defect detection step, the track image is input into a trained neural network model for identifying specific track defects. The model outputs the location information of the identified track defects and converts the location information into the mileage information of the track defects based on the calibration relationship between the two-dimensional camera and the line laser sensor.
7. An automatic track defect location device, employing the automatic track defect location method as described in any one of claims 1-6, characterized in that, include: The detection starting point positioning unit is used to identify the starting point of the track component from the track profile data collected by the online laser sensor based on the profile features of the track component starting point, obtain the corresponding mileage information, and obtain the mileage information of the sleeper based on the mileage information and the predetermined distance information between the starting point of the track component and the nearby sleeper, as the mileage information of the detection starting point. The track defect detection unit is used to detect track defects based on the track profile data or the track image and the mileage information, and to obtain the corresponding mileage information; as well as The track defect location unit is used to obtain at least one sleeper number corresponding to the track defect based on the mileage information of the track defect, the mileage information of the detection starting point, and the sleeper number of the sleeper at the preset detection starting point.
Citation Information
Patent Citations
System and method for inspecting railroad ties
CA2843281A1
Turnout rail piece automatic identification and positioning method and device
CN116923490A
